Introduction: The prognosis for long-term survival even after potentially curative resection for pancreatic adenocarcinoma is poor. Clinical factors determining short-term survival after pancreatic resection are well studied. But prognostic factors predicting long-term survival are poorly understood.
Methods: The records of 362 patients who underwent pancreaticoduodenectomy for pancreatic ductal adenocarcinoma between 1981 and 2001 were retrospectively reviewed and analyzed. Histologic specimens were re-analyzed to confirm diagnosis. Patients with ampullary cancer and patients who underwent total pancreatectomy were excluded. Follow-up was at least 5 years.
Results: In-hospital morbidity and mortality was 39% and 1% respectively. Adjuvant treatment was performed in 77% of patients. Median disease-free survival was 13.8 months with distant and loco-regional recurrence in 64% and 15% of patients respectively. Median survival was 17.4 months. There were 62 five-year survivors (17%), including 21 patients (6%) who survived longer than 10 years. Cohort analysis was performed comparing clinical variables in patients with short-term (<5 years, n=295) and long-term (>=5 years, n=62) survival. Patients’ age and gender were similar in both groups (p>0.05). Univariate analysis showed that more advanced disease (tumor diameter, lymph node stage, ratio of positive to examined lymph nodes) and poor health status (ASA score, serum albumin level) were unfavorable for long-term survival (all p<0.05). Whereas the extent of resection (extended lymphadenectomy, extended pancreatic resection, pylorus-preservation, and portal vein resection) and more aggressive cancers (tumor grade, presence of perivascular or perineural invasion, shorter survival period after diagnosis of recurrence) did not correlate with long-term survival (all p>0.05). Resection margin (p>0.05) was less important than en-bloc resection without penetration of the neoplasm (p=0.005) in order to achieve long-term survival. Multivariate logistic regression analysis identified lymph node stage (OR 0.36, 95% CI 0.14-0.89, p=0.03) as significant prognostic factors for long-term survival. Five-year survival was no guarantee of cure since 16% of this subset died of pancreatic cancer up to 7.8 years after operation.
Conclusion: Operative management of adenocarcinoma in the head of pancreas can be performed with low mortality and acceptable morbidity. Pancreaticoduodenectomy can provide long-term survival in a subset of patients; particularly in the absence of lymph node metastasis. One out of 17 patients can achieve 10-year survival with a potential for cure.